摘要
本文探讨商业银行如何利用贝叶斯分类技术构建企业客户财务危机预测模型。本文使用财务比率作为评价企业绩效的特征属性,并考察两个不同的贝叶斯模型在估计企业客户发生财务危机的后验概率方面的有效性。一个比较简单但有较多的假设,即朴素贝叶斯模型;另一个某种程度上更为复杂但有更少的假设,即组合属性贝叶斯模型。研究发现,与朴素贝叶斯模型相比,由于组合属性贝叶斯模型更好地反映了变量之间潜在的联合分布,因此它能在历史数据支持下估计所要求的概率并做出更精确的预测。所提出的模型可以作为辅助银行审核者做出正确而快速决策的有用工具。
This paper is to demonstrate how to use Bayesian classification technology to build the models of financial distress predictions for corporations in commercial banks. Using financial ratios as predictors of a corporation＇s performance and assessing the posterior probability of a corporation financial health （alternatively, financial distress）, We examine two different probabilistic models, one that is simpler and makes more assumptions, while the other that is somewhat more complex but makes fewer assumptions. The first one is Naive Bayes model, and the second one is composite attributes Bayes model. We find that both models are able to make accurate predictions with the help of historical data to estimate the required probabilities. In particular, the more complex model is found to be very well calibrated in its probability estimates. We posit that such a model can serve as a useful decision aid to an auditor＇s judgment process.
出处
《数理统计与管理》
CSSCI
北大核心
2011年第6期1039-1050,共12页
Journal of Applied Statistics and Management
基金
教育部人文社会科学规划基金项目(项目编号:10YJA910002):信用风险识别的贝叶斯网络模型方法及应用研究
南京理工大学基金培育资助项目(项目编号:2010GJPY020):商业银行贷后信用风险识别模型研究
关键词
分类
信息理论
财务危机
审核
贝叶斯模型
classification, information theory, financial distress, audit, Bayesian models
